Google is building its own DeX: First look at Android's Desktop Mode

This is the thing that bummed me out the most about Microsoft exiting the phone market.

I know Windows isn’t super popular around here, but the idea of carrying one device that I can just dock to work on always intrigued me.

There’s just no way this is taking off with any significant market share in the business world anytime soon being android only, and Apple will never adopt it because they want you to buy 3 different devices. Such a great concept, and with the performance of mobile chips getting so good, very viable.

The problem with this idea is that it's unattainable.

Let's start with something simpler: a living room. There's no universal design that would fit just any living room. The layout and the set of furniture that would work for my living room will not work for yours. Size, shape, windows, doors, connection to other spaces - everything matters. If you want a great design for your living room, you literally need to start from your specific living room.

There's a great idea -- why don't we come with a resizable (reflowable?) design that could fit any living room in the world? While this idea might be entertaining for an engineer's mind, it doesn't work in practice, unless you can settle for just a mediocre design.

Also, being intellectually honest, we need to attack the strongest Apple we can imagine, not a weak Apple that's easy for us to attack. And that strongest Apple will never adopt this idea because they aim to design the best computer/phone/tablet that they can, and in order to design that they need to start with the computer/phone/tablet.

The idea of a phone connecting to a display/keyboard/mouse and becoming a computer has the problem that you could either optimize your design for what you have with a phone, or for what you have with a display and peripherals. It will never be as good as the system designed from the ground up to be just a single thing. It's always nice to have options, but there won't be any mass adoption for the mediocre combo. It was dead in the water with Palm Foleo in 2007, it's just as dead in the water 18 years later.

I'm sorry but I remain unconvinced.

Do you have any examples of desktop UI interfaces that are impossible to create in a Dex like experience, that is possible on Linux, Mac, and Windows desktop experiences?

Or is it that mobile apps will never work great/ideally on desktop? And if that's the case, how is that worse than not having them at all?

Maybe there is a scenario I'm not seeing here.

> could either optimize your design for what you have with a phone, or for what you have with a display and peripherals. It will never be as good as the system designed from the ground up to be just a single thing.

This conclusion looks unsubstantiated to me when you speak of the modern devices that have a sufficient performance for most typical tasks. You probably can't design a gaming computer+phone but nothing prevents you from making a GNU/Linux phone able to work with different desktop environments depending in its current mode. Indeed PureOS and Mobian operating systems already offer that and work well on my smartphone (Librem 5).

I'm not so sure about unattainable. Many programs exist that serve the user in very different contexts, for example responsive websites, web apps. Or games that work on console, like Steam Deck, and PC. Or the Nintendo Switch, which can be used in handheld mode, with a small screen and battery, or docked, connected to a TV. Controllers attached or unattached.

Now, I can see problems too: docked and portable modes need very different performance optimizations. But I'm sure that software can handle this, for example, IDEA IntelliJ has power save mode, and OS-es also demonstrated that they are fine on portable and connected systems alike, like MacOS, Windows, Linux.

It's also not a problem that some things are not available in both modes. For example, Switch has games that explicitly need docked mode, for example, Super Mario Party. Yet both the game, and the platform is popular.

I see no reason why a phone couldn't be a mediocre, or better PC.

While Windows RT / Windows 8 were largely panned, I think they had a lot of potential.

RT / 8 went very hard towards touch interfaces. Then 8.1 made a lot of course corrections to re-establish what's good about mouse + keyboard on Windows.

But... RT was abandoned alongside Windows Phone 7 / 8 and Windows 10 Mobile, and Windows 10 focused on mouse-keyboard once again without nearly as much thought about touch.

I really think this was one case where persistence would have paid off. A focus on "Windows everywhere" instead of (or alongside) "Microsoft apps that all exclusively push cloud services everywhere" could've put Microsoft in a position of mobile and convertible device dominance.

Editing to add, I was just reminded of the Surface Duo -- Android based. (And the announced but abandoned Surface Neo...) Another odd moment I'd describe as... "oh wait let's go back and try our old strategy but without any of the advantages of Windows app support on mobile!"

Not talking about what's possible or impossible, just not seeing it gaining any significant market traction.

Modern pentathlon is a sport where athletes ("pentathletes") compete across five different events. Even the best of them would be mediocre at best at any specific event competing against athletes who have trained for that specific event.

A phone could be a mediocre PC. In my opinion, it will not gain any significant market share competing against other PCs -- mediocre ones, good ones, great ones, and "insanely great" ones too.

> that are impossible to create

My argument is not "this is impossible to create", my argument is "in my opinion this will not succeed in competition against purposefully-made devices." You're welcome to disagree, of course.

> just not seeing it gaining any significant market traction

There can be a lot of reasons for that, like the duopoly forces preventing the competition (https://news.ycombinator.com/item?id=21656355) or insufficient PR.

> Even the best of them would be mediocre at best at any specific event competing against athletes who have trained for that specific event.

In computing, if you have a general-purpose device, it usually can do most tasks sufficiently well.

> duopoly forces preventing the competition

Aka "malice".

> or insufficient PR

Aka "stupidity".

While I love Hanlon's razor as much as anyone, I love Occam's razor even more. And the simple explanation for that is that people don't like to use a mediocre combo when they could be using several purpose-built devices which excel at a specific job.

I like to go around with my Leatherman, but when I have a real job I get my toolbox out instead.

> it usually can do most tasks sufficiently well

An RV could do many "car tasks" sufficiently well, and it can do many "home tasks" sufficiently well, too. And while it has its place, it can't compete against the car + home combo, which is why most people have cars and homes but not RVs.

I don't think we can fairly compare a phone pretending to be a desktop against other desktops.

It would be more fair to compare a phone that has desktop features, to a phone that doesn't have desktop features.

So let's compare the best Apple phone that refuses to have a dex like experience; to a Samsung that has had a dex experience for about 10 years, or to a Google phone that is now adopting desktop experiences.

If the future is anything like the past, in 5 to 10 years from now we'll see a desktop experience on iPhone and they're going to be snobby about it.

> > or insufficient PR > > Aka "stupidity". > > While I love Hanlon's razor as much as anyone

Stupidity isn't necessary here. A small company simply has no resources for a large PR campaign.

Also, nothing prevents a performant iPhone from allowing a full desktop mode except artificial software restrictions resulted from the greediness. Follow the money.

You don't have to degrade the phone experience at all. Just add a Linux VM that it switches to when I plug it into a monitor and has access to my files and we're good.

> I like to go around with my Leatherman

This is exactly what we're asking for. We don't want to get rid of our tablets and laptops. But we want leatherman-level functionality from our phones.

That's perfectly fine. You can get it today from Samsung, or maybe soon from Google. My argument is that Leatherman is not going to sell anywhere remotely to the level of the combination of purpose-made tools that it seems to incorporate. You're welcome to disagree.

> You can get it today from Samsung, or maybe soon from Google.

No, you can't. There's no usable desktop software that you can run there. Artificial restrictions will stay, although might be slightly relaxed.

Most people have a car, not an RV. And when these people chose theirs, was it "fair" comparing a car that cannot sleep 4-6 people to an RV that can?

Well, maybe Librem will make one, and sell it in the billions, who knows.

To me, something along these lines is by far the best approach. Under Android, a Linux desktop could be virtualized on top of Android’s Linux kernel and under iOS, a macOS userland could be virtualized on top of iOS’ Darwin underpinnings.

It’s the only way you don’t end up compromising either half of the experience too much. Trying to converge both into a single UI as Microsoft previously did with Windows and GNOME is now trying to do now is a recipe for failure.

All the plumbing is all there[0].

Here's hoping they hook it up so we can use it.

[0]: https://www.androidpolice.com/android-15-linux-terminal-app/

> There's no universal design that would fit just any living room.

Yes. This is why any furniture store that thinks that they can just offer three designs which would cover 90% of people is going to inevitably fail, as evidenced by Ikea. It's a ridiculous idea that same three designs could be adapted to living rooms across one country, let alone entire globe.

Oh wait. Half of the word's population uses the Bestå cabinets and Nordviken chairs.

I can imagine a scenario when it's really useful, when portability is concerned. This segment right now is served by web apps, which essentially give the same end result, supposing a working internet connection: the user can have the same software and the same files on multiple systems, like a phone, and a PC. The device itself being portable would be the same, but infinitely more private: all the apps and files could live locally. Now, I don't know how large that market is, and I suspect that it's not that large, given that people are just fine with cloud based solutions.

In another words, I think that the functionality itself is very useful. It's just that it's currently being served somewhat adequately by cloud based solutions. For this reason, such a phone-pc product could not offer much in terms of functionality, and so, it might not be popular at all.

It's been already created, albeit it's less powerful than than the mainstream phones. Librem 5 is my daily driver.